The Southern Region
Two winepresses (30, 31
), an arcosolia tomb (204), a bodeda (23) and a cluster of cupmarks (22) were cleared and documented in an area (c. 20×30 m) along the southwestern slopes of the al-Midya hill. The installations yielded no datable finds and on the surface, potsherds dating to the Chalcolithic period, the Early and Middle Bronze Ages, the Iron Age and the Hellenistic, Roman and Byzantine periods were found.
 
Winepress 30 (Figs. 2, 3) was rock-hewn and included a treading floor (2.8×3.6 m) and two collecting vats (northern – 1.1×1.7 m, depth 0.95 m; southern – 0.6×0.9 m, depth 0.9 m).
Three shallow channels led from the floor to the vats, which were linked by a perforated hole in the bedrock. A round sump was cut in each of the collecting vats. An irregular construction (height c. 0.1 m) of small fieldstones and white mortar was close to the southwestern corner of the treading floor; it might have been a repair or a completion of the treading floor side.
 
Winepress 31 (Figs. 4, 5) was hewn in the bedrock slope, c. 20 m above Winepress 30. It included a treading floor (2.6×3.0 m) and a collecting vat (0.9×2.2 m, depth c. 1 m); a shallow channel connected between them. Two levels were hewn in the vat; its eastern part (length c. 0.75 m) was designed as a step with a circular sump in the corner, which was c. 0.1 m higher than the western part of the vat.
 
Four circular cupmarks (22; diam. 0.2–0.3 m) and a shallow bodeda (23; 0.5×0.8 m; Figs. 4, 6) with a small depression (diam. 0.15 m) hewn in its northwestern corner were cut in the bedrock, c. 20 m north of Winepress 31. A rectangular opening of a tomb (204) was identified c. 20 m to the north; although it was not excavated, the end of a curved ceiling of two arcosolia could be discerned inside (Figs. 7, 8).

The Northern Region
On the surface at the northwestern foot of the al-Midya hill, on the western side of the stream, c. 10 stone clearance heaps were observed. Mechanical equipment trenched the stone clearance heaps, in the wake of which the heaps were removed together with a layer of alluvium that covered the bedrock surface (c. 30×50 m). A concentration of rock-hewn installations was discovered, including three clusters of cupmarks (33, 35, 36); single cupmarks scattered throughout the area, three of which were documented (16, 18, 20; diam. 0.8–1.0 m, depth 0.5–0.6 m; Fig. 9); two hewn circles (34, 37); a cistern (32); three pits (15, 17, 40); a bodeda (41) and a natural hollow (19; diam. c. 2 m, max. depth 1.1 m; Fig. 10) that may have been used. Two sections of walls (14) were discovered c. 50 m to the southwest. A cave (7) was inspected further to the south, on a precipitous slope above the stream.

Cupmarks. Clusters of cupmarks (33, 35, 36), hewn on bedrock surfaces close to each other, extended across an area of c. 6.5×8.0 m (Fig. 11). Cluster 35, in the eastern part of the bedrock surface, consisted of fourteen small round cupmarks (diam. c. 0.1 m) in two groups. Six elliptical cupmarks (max. diam. c. 0.3 m; Fig. 12) were hewn in Cluster 33; they resembled cupmarks found in Modi‘in, which were identified as installations for pounding and threshing that dated to the Chalcolithic period (HA-ESI 119). A cluster of six round cupmarks (36; diam. 0.1–0.2 m; Fig. 13) was discerned c. 5.5 m to the west. Small worn body fragments that could be attributed to pottery vessels from the Early Bronze Age or the Chalcolithic period were found in most of the cupmarks. The date of the cupmarks could not be determined with certainty; however, it seems that the meager finds indicated they probably dated to one of these two periods.
 
 
Cistern 32 (depth c. 2.5 m; see Fig. 11) was hewn in the southern part of the bedrock surface where the cupmark clusters were found. The elliptical opening (c. 0.7×1.0 m; Fig. 14) of the cistern was positioned in a manner that allowed a natural bedrock fissure (length c. 6 m, width 0.1 m) to convey water into the cistern from the northeast. A small depression (diam. c. 0.7 m, depth 0.3 m) was hewn in the bottom of the cistern, which yielded only a few non-diagnostic potsherds.
 
Pits. The pits had a round or circular opening and a circular outline (15 – diam. 1.5 m, depth 1.8 m; 17 – diam. 1.3–1.7 m; Figs. 15, 16). Pit 15 had a flat bottom, whereas the excavation in Pit 17 was 2.8 m deep, yet did not reach its bottom. Several worn and non-diagnostic potsherds were found in the pits. Pit 40 (diam. 0.8 m, depth 1 m; Fig. 17) was probably meant to be larger and deeper, similar to Pits 15 and 17, but the quarrymen did not complete the rock-cutting, as evidenced by the sides that remained jagged; it is possible that the quarrymen encountered hard bedrock in this spot. A Canaanean sickle blade, retouched on its distal end, delicately denticulated along its cutting edge and dating to the Early Bronze Age (Fig. 18), was found inside the pit. Two rock-cut circles (34, 37; diam. c. 1.8 m), which possibly marked the outline of non-hewn pits, were discerned 0.5–1.5 m to the northwest.

The Bodeda (41; Fig. 19) was hewn alongside Cupmark Cluster 33. It consisted of a small cupmark and a shallow square treading surface (0.2×0.4 m), in whose southeastern side was a settling depression (diam. 0.15 m).
 
Wall Sections. Southwest of the rock-hewn installations on the surface, stones protruding from two walls could be discerned; the walls were partially exposed in the excavation (14; Fig. 20). Close to the walls (W127 – overall length 7 m, width 1.2 m; W128 – overall length 2.5 m, width 1.1 m), which were founded directly on the bedrock, were fragments of a bowl (Fig. 21:1) and a krater (Fig. 21:2), dating to the Early Bronze Age. As no floor was discovered, it seems that these were retaining walls in the corner of a farming terrace.

Cave 7 (Fig. 22) was located on the western slope of Nahal Modi‘im, which was difficult to reach; however, a trench dug in the slope for the road construction, facilitated the access to it. A stone wall (W11; Fig. 23) blocked the entrance of the cave and was possibly built in the Mamluk period, as evidenced by a silver dirham from the time of
Al-Adel Zein Al-Din Kitbugha (1295 CE; IAA 85691) that was found in the cave’s opening. The excavation in the front portion of the cave (2.3×5.0 m, height 1.3 m), beneath lumps of bedrock that had collapsed from the ceiling (L1) and a layer of brown alluvium (L2), revealed a level of ceramic finds (L3) that included a cooking pot (Fig. 24:1), jars (Fig. 24:2–5) and juglets (Fig. 24:8, 9) from the Iron Age, as well as jars (Fig. 24:6, 7) and a lamp (Fig. 24:10) from the Late Hellenistic and the beginning of the Roman periods. A narrow corridor (length c. 2.5 m, width 1.3 m) that split into two passages (width 0.8–1.0 m) extended back beyond the front space of the cave (length c. 6 m, max. width 2.3 m). The northern passage (L120) terminated after c. 2 m; the southern passage (L121) was examined for a distance of c. 5 m, without locating its end. The accumulation on the bottom of the southern passage yielded a silver Roman dinar from the time of Hadrian (119–138 CE; IAA 85653; Fig. 25). Further along the southern passage was a fragment of a lamp dating to the Islamic period (Fig. 24:11). The finds from the cave indicate that despite the difficult access to the cave, it was used over a prolonged period of time and may have served as a hiding refuge during the Bar Kokhba Revolt.